UResearch

Inspiration

We realized that students at the University of Rochester often struggle to discover relevant research opportunities, papers, and faculty work. Information is scattered across departmental websites, outdated pages, or word of mouth. Professors also lack a centralized, student-friendly space to share their research. We wanted to build an integrated, intelligent platform that connects students with faculty research instantly using semantic search-something fast, intuitive, and aligned with the UR community.

What it does

UResearch is a unified research discovery platform where professors can post their research articles, summaries, and updates. Students can then use our semantic search engine to explore related topics, find relevant papers, and connect with research directions they care about. Instead of keyword matching, the platform understands meaning-making it easier for students to find what they actually need.

How we built it

We built UResearch using: Next.js for a fast, scalable frontend. TailwindCSS for UI. Vector embeddings + semantic search to match user queries with research posts. A structured posting portal for professors to upload articles with metadata.

Challenges we ran into

Designing a clean, content-focused UI while keeping UR branding consistent. Structuring research data so semantic search could classify and retrieve it accurately. Ensuring search results remained relevant even with limited initial data. Balancing simplicity for students with the advanced tools professors expect. Avoiding breaking existing code as we kept improving the UI.

Accomplishments that we're proud of

Building a full end-to-end semantic search experience in a short timeframe. Creating a professional, modern UI that feels like an official UR system. Providing a platform that genuinely solves a real problem for students and professors.

What we learned

How to integrate vector search into a real-world interface. The importance of UI clarity: students engage more with clean, compact layouts. The value of small, modular components in Next.js when iterating quickly.

What's next for UResearch

Adding professor profiles with research interests and current lab opportunities. Implementing filters (fields, tags, departments) to refine semantic search further. Allowing students to bookmark, follow labs, and get notifications for new posts. Expanding semantic search to support PDF uploads and automatic article summarization. Scaling the platform to all research groups across the University of Rochester.